硕士学位论文
响沙湾推荐系统对网络消费者购买行为的
25岁属什么生肖影响研究
IMPACT OF RECOMMENDER SYSTEMS ON ONLINE CONSUMER PURCHASING BEHA VIOR
孙晓梅
哈尔滨工业大学
2011年12月
国内图书分类号：F713.36 学校代码：10213 国际图书分类号：004.3 密级：公开
管理学硕士学位论文
推荐系统对网络消费者购买行为的
晨跑的好处影响研究
硕士研究生：孙晓梅
导 师：杜桂荣副教授
申请学位：管理学硕士
学科：管理科学与工程
所在单位：深圳研究生院
答辩日期：2011年12月
形容山水的词语授予学位单位：哈尔滨工业大学
Classified Index: F713.36
U.D.C: 004.3
Thesis for the Master Degree in Management
IMPACT OF RECOMMENDER SYSTEMS ON ONLINE CONSUMER PURCHASING BEHA VIOR
Candidate：Sun Xiaomei
Supervisor：Prof. Du Guirong
Academic Degree Applied for：Master of Management Speciality：Management Science and Engineering Affiliation：Shenzhen Graduate School
铁甲骑士2中文版
Date of Defence：December, 2011
Degree-Conferring-Institution：Harbin Institute of Technology
嫉恨的读音>决策力哈尔滨工业大学管理学硕士学位论文
摘要
冬季训练随着电子商务的蓬勃发展，越来越多的商家建立网络商店以向消费者出售产品与服务，消费者在选择增多的同时过量的产品信息也给其带来了信息过载等问题。为减轻此类问题，几乎所有大型的电子商务网站，都在一定程度上使用各种形式的推荐系统服务，以帮助消费者更加便利地搜索产品，并根据消费者的需求偏好做出推荐，为消费者提供个性化的“导购”服务。为了研究消费者是否会使用推荐系统的推荐结果来做出购买决策，本文在对推荐系统与网络购物的相关研究进行回顾梳理的基础上，探
讨了推荐系统对网络消费者购买行为的影响，以认知信任和情感信任为中心视角，研究感知定制化、熟悉度、来源可信度对消费者采纳推荐信息意愿以做出购买决策的影响作用。研究目的是为了更好地设计推荐系统，使消费者对其更加信任并采取推荐信息做出购买决策。
本研究按标准的方法步骤进行了问卷调查，利用SPSS16.0和AMOS16.0统计软件对样本数据进行数据统计分析，对各变量进行描述性统计，评估了感知定制化、熟悉度、来源可信度等测量量表的信度，通过相关性检测了各变量之间的相关度，通过验证性因子分析方法检验研究模型中各变量之间的区分效度和模型的匹配度，最后，利用全模型分析来检验本文提出的研究假设。
结果发现，感知定制化、熟悉度均正向影响消费者对推荐系统的认知信任；来源可信度正向影响消费者对推荐系统的感知正直信任；认知信任正向影响情感信任；情感信任正向影响消费者采纳推荐信息的意向。
关键词：推荐系统；认知信任；情感信任；采纳意向
哈尔滨工业大学管理学硕士学位论文
Abstract
With the booming development of electronic commerce, more and more online stores are t up to s
ell products and rvices to consumers, which provides consumers with more choices, however, excessive product informaiton brings problems as well, such as information overload. In order to decrea this problem, almost all large e-commerce companies provide various forms of rvices from recommender system. Recommender system technology offers an effective solution for managing information overload and complexity in e-commerce. Recommender system is a personalized computer agent that gives an online customer advice about what to buy bad on the customer’s personal needs and preference. Recommendation information adoption is studied with a trust-centered, cognitive and emotional balanced orientation. This study theoretically articulates and empirically examines the effects of perceived personalization, familiarity and source credibilit on cognitive trust and emotional trust in a recommender system, and the impact of cognitive trust and emotional trust on the intention to adopt the recommendation information. The purpo of the study focus on how a recommender system can be designed better so that a customer will be more likely to trust and then adopt the recommender system for the customer’s purcha decision making.
A standard questionnaire survey is conducted and the analysis of the sample data is proceeded in SPSS16.0 and AMOS 16.0. First, the descriptive statistics for each variable are made and the scale reliability are evaluated. Then, the correlation analysis of the variables are procesd as well as conf
irmatory factor analysis. Finally, the full model analysis is ud to test the hypothesis propod in this study.
The primary empirical results shown that perceived customization and familiarity positively influence cognitive trust in recommender systems, source credibility positively influences consumer perception of integrity beliefs of the recommended system, cognitive trust positively effects emotional trust, emotional trust in the recommender system significantly increas a customer’s intention to adopt a recommendation information.
Keywords: recommender system, cognitive trust, emotional trust, adoption intention
